4910
J. Org. Chem. 1983,48,4910-4914
previously4 as were donors 3-5.3 Decompositions were carried out under conditions described in Table I, following peroxide decomposition by iodimetry. Reactions in the presence of 4 were too rapid for accurate measurement, while for 5, an experiment in an NMR tube a t -20 "C showed complete reaction in 5,i t is well approximated (within 10%)by the steady-state assumption, giving k 2 / k l = [A]/[B], regardless of how much A actually goes to D.
Walling, Zhao, and El-Taliawi hol/ toluene can also be calculated from the date of Table V and of earlier tables where complete material balances are r e p ~ r t e d .Results ~ are not very accurate since benzyl alcohol yields are low but runs 2-4 in Table I11 give values for k (benzyl alcohol): k (toluene) of 19 f 4 (standard deviation of the mean) in acetonitrile while Table I1 indicates slightly lower values in acetic acid: -12 (run 1) and 7 (run 2 with added acetate). The experiments in Table V in acetonitrile indicate an increase in selectivity in the order p-OCH, < p-CH3 < p-C1 (19, 67, and 112, respectively), and those in acetic acid show a similar trend. Although the relative reactivities of substituted toluenes parallel relative rates of reaction of substituted benzoic acids with SO4-. as measured directly in water (3.5 x lo9, 1.8 X lo9, and 0.36 X lo9 L mol-' s-l for p-OCH3, p-CH3, and p-Cl)15and are thus consistent with eq 2, SO4--attack on substrate, being the rate-determining step in competitive oxidations, the high reactivities of benzyl alcohol compared with those of toluenes appears inconsistent with this. Since the alcohols contain an electronegative oxygen, this should, if anything, decrease the ease of formation of the resulting radical cation, and the implied reaction rates (lO1o-lO1l mol L-' s-l) are too high to be plausible. Two alternative explanations appear possible. Radical cations could be partially equilibrated by electron transfer, i.e., eq 6. Since proton loss from benzyl alcohol radical cations At*
t 8
A f
Et*
I
1
products
products
(6)
is apparently much faster than from toluene radical cations,16this would account for their selective oxidation. A similar suggestion has been made by Minisci to account for the selective oxidation of p-methoxytoluene in a mixture with the meta isomer.6 Alternatively, in these nonaqueous media, Cu could be selectively solvated by the alcohols. As SO4-. was generated by reaction 5, it would be in immediate proximity to the benzyl alcohol and could react with it within the solvent cage. Such selective solvation of Fe in acetonitrile has been observed by Groves,17 and there is evidence for such cage reactions in the Fenton's reagent oxidation of mandelic acid in water.18J9 Oxidation of Other Alkylbenzenes. Finally, we have carried out some preliminary experiments on the oxidation of benzenes with more complex alkyl side chains (Table VI). With ethylbenzene the results are straightforward and resemble those with toluene, except that now Cu oxidation of the intermediate benzylic radical yields some styrene, a well recognized path in the oxidation of radicals with @-hydrogenin these media.20 With isopropylbenzene, conversion to a-methylstyrene is evidently the major reaction path since it is the major product in reactions run for shorter times. Acetophenone and 2-phenylpropanal evidently come from its further oxidation, since they are obtained in high yield from ox(15)Neta, P.;Madhavan, V.; Zemel, H.; Fessenden, R. W. J. Am. Chem. SOC.1977,99,163. (16)The differences in rate in water have been estimated as over lo3: Walling C.; Camaioni, D. M.; Kim, S. S. J. Am. Chem. SOC.1978,100, 4814. (17)Groves, J. T.; Van Der Puy, M. J.Am. Chem. SOC.1976,98,5290. (18)Walling, C.; Amarnath, K.; Campbell, C. B. J. Am. Chem. SOC. 1982,104,1185. (19)A third possibility, concerted oxidation and proton loss in benzylic alcohols, has been suggested in water4 but now seems unlikely.16 In any case, it can hardly account for the implausibly high rates for the SO4-. -alcohol rates required by the data. (20)Kochi, J. K.; Bemis, A.;Jenkins, C. L.J. Am. Chem. SOC.1968, 90, 4038,4616.
J. Org. Chem., Vol. 48, No. 25, 1983
Oxidation of Alkylbenzenes by szo82--cu"
4913
Table V. Competitive Oxidations run A'" Bb
% yieldC
toluene substituent
% conversion
ArCHO
ArCH,OAc
ArCH,OH
p-OCH, P-CH, p-c1 p-OCH, P-CH, p-c1
50 18 10 40 13 9
16 17 14 67 67 73
70 76 80 4 5 4
5 4 trace 8 10 9
a 2 mmol of each toluene, 4 mmol of (NH,),S,O,, 1 mmol of Cu(OAc), in 5 mL of HOAc, and 1 mL of H,O, a t 120 "C for 3 h. Same as in a but in 5 mL of acetonitrile and 1 mL of H,O a t 100 "C for 3 h. Based on toluene consumed.
Table VI. Oxidation of Alkylbenzenes substrate
conditionsa
% conversion
ethylbenzene
HOAc
62
ethyl benzene
acetonitrile
95
cumene
HOAc
33
cumene
acetonitrile
22
p-ethyltoluene
HOAc
66
p-ethyltoluene
acetonitrile
24 (28)
p-cymene
HOAc
61
products 1-phenylethyl acetate acetophenone 1-phenylethanol acetophenone 1-phenylethanol styrene a-methylstyrene acetophenone 2-phenylpropanal 1-phenylethyl acetate cumyl alcohol a-methylstyrene acetophenone 2-phenylpropanal 1-phenylethyl acetate cumyl alcohol p-methylstyrene 1-p-tolylethanol p-ethyl benzaldehyde p-methylacetophenone p-ethylbenzyl acetate p-methylstyrene 1-p-tolylethanol p-ethylbenzaldehyde p-methylacetophenone p-ethylbenzyl acetate p , a -dimethylstyrene 2-p-tol ylpropanal
p-cymene
acetonitrile
12
p-isopropylbenzaldehyde p-methylacetophenone p-isopropylbenzyl acetate pia -dimethylstyrene 2-p-tolylpropanal p-isopropylbenzaldehyde p-methylacetophenone p-isopropylbenzyl acetate
% yield b , c
60 33 5 88
5 7 21 25 47 3.6 3.6 69 21 10 tr tr 13 (10) 0.8 ( 2 7 ) 8.3 ( 2 ) 43 (36) 35 (24) 22 (37) 3 (1) 21 (6) 52 ( 2 6 ) 3 (31) 1.5 23 12 9 54 26 4 39 32 tr I
2 mmol of substrate, 4 mmol of (NH,),S,O,, 1 mmol of Cu(OAc), in 5 mL of solvent, and 1 mL of H,O a t 100 "C. Based on substrate reacted. Numbers in parentheses are for reactions with 4.5 mmol of NaOAc added.
a
idation of a-methylstyrene itself.21 Again, this seems to be a case in which the initial product, a-methylstyrene, is oxidized more rapidly than the starting material, cumene, by a factor of 5-10 if we treat the data in the same way as those for the benzyl alcohol-toluene system. Finally, oxidation of p-ethyltoluene and p-cymene yield products indicating competing loss of protons from both alkyl groups of the intermediate radical cation. There has been an impression that oxidation of methyl occurs preferentially or exclusively;6~22~23 for example, p-isopropylbenzyl acetate has been reported as the sole isolated product (70%)from p-cymene oxidation by s2082--cu in acetic acid.22 However, under our conditions this is clearly (21) This interesting reaction is being examined further and will be reported elsewhere. (22) Onopchenko, A.; Schulz, J. G. D.; Seekircher, R. J. Org. Chem. 1972, 37, 1414, 2564. (23) Giordiano, C.; Belli, A.; Citterio, A.; Minisci, F. J. Org. Chem. 1979,44, 2314-5.
not the case. From the data, corrected statistically for the number of protons available, the ease of loss of primary/secondary/tertiary protons is about 1:2:1.6in acetic acid. In acetonitrile the selectivity is greater, 1:4.8:4.8. On the basis of radical stability, an order of primary < secondary < tertiary might be expected, and the numbers may indicate some steric hinderance in the tertiary case.
Summary (1) Our results confirm that S,O:--Cu" oxidation in acetic acid and acetonitrile is a promising technique for generating aromatic radical cations and studying their subsequent reactions. Although some SzO:- is lost by oxidation of solvent, identifiable products are obtained in high yield based on substrate consumed. (2) Not only are toluenes converted to benzyl acetates in acetic acid as previously reported but also the reaction in acetonitrile gives benzaldheydes in high yield, so both reactions appear synthetically useful.
J. Org. C h e m . 1983, 48, 4914-4917
4914
(3) Ethylbenzene and cumene are also readily oxidized, cumene giving initially principally a-methylstyrene which is rapidly further oxidized to 2-phenylpropanal and acetophenone. (4) p-Ethyltoluene and p-cymene undergo competitive oxidation of both alkyl groups with a 2-4:l preference for proton loss from the more substituted benzylic carbon. (5) In competitive oxidations, toluenes with electronsupplying groups are attacked preferentially. Similarly, benzyl alcohols and styrenes are oxidized in preference to their parent alkylbenzenes. Here the large spread in reactivity is inconsistent with known or expected rates of SO4-. reactions with these substrates and may arise from electron transfer between radical cations and substrates, with the result that relative rates of side-chain cleavage become important in determining apparent relative substrate reactivities. In fact, the variation in product distributions under different conditions which we find and discrepancies between our findings and those in the literature imply that radical cation fragmentations are quite complex and will reward further study.
Experimental Section Reagents and solvents were commercial materials, and purity checks were by GLC or other appropriate means. Reactions were in general carried out in sealed 15-mL Teflon-capped vials flushed with argon or nitrogen and magnetically stirred in thermostated baths as indicated in the tables. In runs
where only major products were determined, reaction mixtures were analyzed directly by GLC against internal standards (Carbowax OV-17 column) and comparison with authentic material. When complex mixtures of products were obtained, e.g., Table VI, reaction mixtures were poured into water, extracted with ether, dried, and, after removal of ether, analyzed by a combination of GLC and GC/MS (VG Micromass 7070 mass spectrometer)with comparison with reference spectra when available. Thus 2phenylpropanal [M+.,m l e 134; base peak, m l e 105 (M - CHO)] matched the reference spectra for authentic material, and 2-ptolylpropanal (M+.,m / e 148, base peak, m / e 119 (M - CHO); also m / e 91 (C7H7+), and a general fragmentation pattern similar to 2-phenylpropanal above] was clearly distinguished from p-isopropylbenzaldehyde [M'., m / e 148;base peak, m l e 133 (M - CH,); also m l e 119 (M - CHO) and 105 (M - C3H7)].
Acknowledgment. Support of this work by a grant from the National Science Foundation is gratefully acknowledged. GC/MS analyses were obtained from our mass spectra laboratory, support of which by NSF Grant CHE-8100424 and the University of Utah Institutional Funds is also acknowledged, and we thank Dr. Thomas Sharp of that laboratory for his assistance in interpretation of the results. Registry No. p-Methoxytoluene,104-93-8;p-methyltoluene, 106-42-3; p-chlorotoluene, 106-43-4; ethylbenzene, 100-41-4; cumene, 98-82-8; p-ethyltoluene, 622-96-8;p-cymene, 99-87-6; toluene, 108-88-3; (NH&S208, 7727-54-0; KzSz08, 7727-21-1; CU(OAC),,142-71-2; CBH,C(CH,)=CHz, 98-83-9.
Oxidation of Arylalkanols by S202--Cu" Cheves Walling,* Gamil M. El-Taliawi,' and Chengxue Zhao2 Department
of
Chemistry, University of Utah, Salt Lake City, Utah 84112 Received J u n e 1, 1983
Products of oxidation of a series of arylalkanols by Sz082--Cu"in acetic acid and acetonitrile are consistent with initial oxidation to aryl radical cations followed by either loss of a benzylic proton or C-C bond scission and subsequent oxidation of intermediate benzylic radicals by Cun to final products. 1-Arylalkanols(ArCHOHR) react by both paths, C-C bond scission increasing with stability of the radical R.. 2-Arylalkanols (ArCH,CHOHR) give chiefly C-C bond scission to benzyl radicals and RCHO. The 3- and 4-arylalkanols undergo chiefly proton loss, and the resulting radicals are oxidized with cyclization to chromans and 2-phenylfurans,respectively, both of which may be oxidized in turn to further products.
It is generally accepted that side-chain oxidations of aromatic molecules by higher valence metals and other reagents which may act as one-electron oxidants often involve initial oxidation of the aromatic to a radical cation.= The radical cation then undergoes loss of a benzylic proton or the side-chain fragmentation, and the resulting radicals are further oxidized to final products. The use of peroxydisulfate plus Cu(I1) to oxidize intermediate radicals is a convenient technique for studying such reactions, and we have recently described the effect of reaction variables in the oxidation of toluenes and some other alkylbenzenes in acetic acid and a~etonitrile.~Of particular interest are molecules where competing paths of fragmentation are available. Here an interesting example (1) Visiting Associate Professor, University of Cairo, Cairo, Egypt. (2) Visiting Scientist, Shanghai Institute of Organic Chemistry, Shanghai, People's Republic of China. (3) Norman, R. 0. C.; Storey, P. M. J. Chem. SOC.E 1970, 1099. (4) Snook, M. E.; Hamilton, G. A. J. Am. Chem. SOC.1974, 96,860. (5) Walling, C. Acc. Chem. Res. 1975, 8, 125. (6) Minisci, F.; Citterio, A.; Giordano, C. Acc. Chem.Res. 1983,16, 27. (7) Walling, C.; Zhao, C.; El-Taliawi, G. M. J. Org. Chem., previous paper in this issue.
0022-3263/83/1948-4914$01.50/0
Scheme I OH
OH
I JH-R
A
,C-R
0
II
r-D
OH*
I/
dH - CYCH0 has been reported by Snook and Hamilton: who found that Sz082- oxidation of arylalkyl carbinols (ArCHOHR) in aqueous solution occurs by two paths involving the competing cleavages (Scheme I). As might be expected, path B was favored by increasing the stability of the radical R. in the order Me < Et < i-Pr < t-Bu. We have now examined the competition in acetic acid and acetonitrile and find that the competition is also 0 1983 American Chemical Society
